2 BULLETIN OF THE 



ous to get an idea of its mode of formation, because, according to Dar- 

 win's theory of coral reefs, such atols should not occur in areas of 

 elevation like those in which the Florida reefs, the coast reefs of Cuba, 

 the Bahamas and the Central American reefs are found. The examina- 

 tion of Alacran Reef showed it to be in full activity, the eastern slope 

 being nearly perpendicular, rising in a comparatively very short distance 

 from 20 fathoms to the surface. The eastern side forms a huge semi- 

 circle exposed to the full force of the northeast trades, and the surf 

 breaks heavily against the great masses of Madrepora palmata which 

 build up the narrow line of coral just awash or barely flush with the 

 level of the sea. The greatest length of the reef is about fourteen 

 miles ; its width, eight miles. It forms a very regular elliptical ring, the 

 eastern half of which rises boldly from 20 fathoms, while the western half 

 consists of half-finished narrow islands. These islands are mere strips of 

 sand formed by the breaking up of the exposed coral heads, which are 

 gradually cemented together by the accumulation of the loose materials, 

 sand, etc., held in suspension by the water, and constantly driven west- 

 ward by the northeast trades. The space intervening between the two 

 edges of the reef is from one to six fathoms in depth, filled in its shallower 

 parts by huge masses of Astrroa, of Gorgonife, of Maeandrina, of Madre- 

 pora palmata, which occasionally rise to the surface. The intervening 

 spaces become choked by the sands formed by the decomposition of the 

 eastern edge of the reef, under the unceasing action of the heavy surf. 

 These heads are thus separated by lanes of sand, forming deep gullies in 

 which the growth of the corals or Gorgonia; has been arrested by the 

 prevailing drift to the west. These heads are gradually but steadily 

 approaching the western strip of islands, and will in time completely fill 

 the whole space in the interior of the reef, or will leave only a narrow 

 channel varying in depth between the outer broad bank to the east and 

 the narrow islands to the west. The corals on the west side of the 

 western islands, and also on the western slope of the reef, grow on a 

 much more gentle slope, the 10-fathom curve being from an eighth to half 

 a mile or even more distant from the shore of the islands. Other reefs are 

 beginning to rise on the Yucatan Bank, near the 30 or 20 fathom curve. 

 The whole structure of this reef shows its identity of formation with 

 that of the main Florida Reef, and with that of the reefs on the north- 

 ern coast of Cuba, where the line of distinct and powerful elevation can 

 be still plainly traced by old coral slopes and by the ancient coral reefs 

 in the hills surrounding Havana and extending to Matanzas. These hills 

 attain a height of over 1,200 feet, and arc entirely composed of species 



